Blatant fraud in action.
-
ict558 wrote:
"If people thought that the GISS approach was deeply flawed, they are deeply wrong."
Lets look at Hansens methods are: 1) UHI adjustments using night time satellite images showing artificial light density. While this is an indication of the degree a region is urbanised it does not answer the positioning of individual stations. If a station is badly sited locally, ie above concrete or tarmac. Or near a source of heat, then the density of street lights in an area a few square miles around it is not going to show that. 2) Hansen uses a 1500 mile smothing algorithym so that the temperatures of a particular station get applied to a vast area. If you look at his colour temperature map you might ask yourself why Greenland is 5 degrees hotter. The answer is because the station data applied to the 15000 mile region comes from a station at an airport. A station in a low density area but regularly impacted by jet exhaust. 3) Station selection has also caused alarm. Over the last 20 years less and less stations are used to compose the data set, favouring lower altitude and lower lattitude stations and stations in urbanised environments, such as airports. A number of papers have been published which attempt to show that this is of no concern. However it does leave a question as to why stations have been deselected. When you add all that together I have very little faith in his product. Tell me, have you ever looked at raw station data from around the world?
============================== Nothing to say.
This thread is "Blatant fraud in action".
Erudite_Eric wrote:
Lets look at Hansens methods are:
Nothing here to suggests blatant fraud. Each point is discussed in detail by GISS, and their reason for adopting the processes they do is explained. Bernie Madoff was a fraudster, nobody knew what he was doing with the money he was given. GISS are not fraudsters, anybody can find out what they are doing with the temperatures they are given. Points: 1. Which is why other proxies for 'urbanization' are currently being investigated. E.G.: The Impact of Urbanization on Land Temperature Trends Zeke Hausfather, Steven Mosher, Matthew Menne (NCDC NOAA), Claude Williams (NCDC NOAA), and Nick Stokes. 2. GISS provides two radii of influence, 250km and 1200km (the standard analysis). "... it is possible that the GISS analysis overstates the magnitude of Arctic warming in regions where data are extrapolated ..." 3. "GISS uses all of the GHCN stations that are available, but the number of reporting meteorological stations in 2009 was only 2490, compared to ∼6300 usable stations in the entire 130 year GHCN record." So they weren't deselected.
Erudite_Eric wrote:
If you look at his colour temperature map you might ask yourself why Greenland is 5 degrees hotter.
Anomalies/Trends, Period, F/C, hotter than what?
If people made the effort to read something three times before commenting, blogs would be much more useful places. - Anon.
-
Erudite_Eric wrote:
I just pointed out to you how you can clearly see how his method is not good enough
That wasn't in this sub thread. It wasn't in the link and it wasn't in your first comment either.
Erudite_Eric wrote:
I then poiint out Hansen uses night time satellite images to adjust for UHI and sugfgest you go to surfacestations.org to see how stations can be badly sited in an area of very low density street lighting and you say:
That by itself doesn't seem convincing to me. But perhaps you actually expanded on that, measured the problems with stations (not hypothetical) as a statistical variance and then applied that to the existing data.
jschell wrote:
But perhaps you actually expanded on that, measured the problems with stations (not hypothetical) as a statistical variance and then applied that to the existing data
Yeah, sure. No, what I did was look at how individual stations are sited and decide that adjusting for that using satellite images is insuficient. In stead urbanised stations should be rejected, and only rural ones kept in the data set.
============================== Nothing to say.
-
This thread is "Blatant fraud in action".
Erudite_Eric wrote:
Lets look at Hansens methods are:
Nothing here to suggests blatant fraud. Each point is discussed in detail by GISS, and their reason for adopting the processes they do is explained. Bernie Madoff was a fraudster, nobody knew what he was doing with the money he was given. GISS are not fraudsters, anybody can find out what they are doing with the temperatures they are given. Points: 1. Which is why other proxies for 'urbanization' are currently being investigated. E.G.: The Impact of Urbanization on Land Temperature Trends Zeke Hausfather, Steven Mosher, Matthew Menne (NCDC NOAA), Claude Williams (NCDC NOAA), and Nick Stokes. 2. GISS provides two radii of influence, 250km and 1200km (the standard analysis). "... it is possible that the GISS analysis overstates the magnitude of Arctic warming in regions where data are extrapolated ..." 3. "GISS uses all of the GHCN stations that are available, but the number of reporting meteorological stations in 2009 was only 2490, compared to ∼6300 usable stations in the entire 130 year GHCN record." So they weren't deselected.
Erudite_Eric wrote:
If you look at his colour temperature map you might ask yourself why Greenland is 5 degrees hotter.
Anomalies/Trends, Period, F/C, hotter than what?
If people made the effort to read something three times before commenting, blogs would be much more useful places. - Anon.
Oh stop being so reasonable! :mad: You are destroying a perfectly good rant! :mad: ;P Who was it who finally squeezed Hansens data and methods out of him a few years back and pointed out a massive error in his work that actually ended up with the 30's being hotter than the 90's? I forget who it was, but Hansen has been anything BUT open and honest. As for reporting stations, I am not sure I believe you. If this is such an important issue how can station data be left unreported? The scientific world wil surely extract some more of the tax payers loot in order to make sure station data is forthcoming. As for greenland, its the anomoly. An anomoly derived fron the temperature sensor at the end of run way 12, right where the jet trubines get cranked up. :)
============================== Nothing to say.
-
jschell wrote:
But perhaps you actually expanded on that, measured the problems with stations (not hypothetical) as a statistical variance and then applied that to the existing data
Yeah, sure. No, what I did was look at how individual stations are sited and decide that adjusting for that using satellite images is insuficient. In stead urbanised stations should be rejected, and only rural ones kept in the data set.
============================== Nothing to say.
Erudite_Eric wrote:
Yeah, sure.
Not sure what you mean. What I proposed fits within the scientific methodology.
Erudite_Eric wrote:
No, what I did was look at how individual stations are sited and decide that adjusting for that using satellite images is insuficient. In stead urbanised stations should be rejected, and only rural ones kept in the data set.
That is a statement of an unproven theory. It seems provable but also must account for showing that once the data is massaged that way that it still represents a reasonable representation of actual weather. But until proven it does not refute the original link.
-
Oh stop being so reasonable! :mad: You are destroying a perfectly good rant! :mad: ;P Who was it who finally squeezed Hansens data and methods out of him a few years back and pointed out a massive error in his work that actually ended up with the 30's being hotter than the 90's? I forget who it was, but Hansen has been anything BUT open and honest. As for reporting stations, I am not sure I believe you. If this is such an important issue how can station data be left unreported? The scientific world wil surely extract some more of the tax payers loot in order to make sure station data is forthcoming. As for greenland, its the anomoly. An anomoly derived fron the temperature sensor at the end of run way 12, right where the jet trubines get cranked up. :)
============================== Nothing to say.
Erudite_Eric wrote:
Who was it who finally squeezed Hansens data and methods out of him a few years back
Strangely, the data processing method has always been available. However, the NCDC data were formatted for processing, i.e. binary data. The resultant GISS data were also held as binary files. I do remember reading the squeals when the binary files and FORTRAN programs were made available to the 'skeptic' community. (Huh! Amateurs! :-D ) Funny how the Open Source guys rewriting the FORTRAN code in Python and creating text data files had no problems communicating with GISS. Perhaps it was because their attitude was reasonable ( ;P ) rather than confrontational.
Erudite_Eric wrote:
a massive error in his work that actually ended up with the 30's being hotter than the 90's
No error in his work, rather a massive error in the nature of the US station data being passed to GISS by NCDC between 2000 and 2007. This affected US temperature anomalies from 2000. (Nothing to do with the 30's and 90's.) When the data were corrected in 2007, the error was shown to have had no effect on the Global Land-Ocean anomalies.
Erudite_Eric wrote:
As for reporting stations, I am not sure I believe you.
E.G., Canada. GHCN station count: 700-800 in the 1960's, 500-600 in the 1980's, -400 for 1990, <100 for 1991, ~50 thereafter. Although there are 1480 stations in Canada, only 49 are reporting to GHCN via the automated CLIMAT system. So the drop-out appears to be due to automated data collection.
Erudite_Eric wrote:
If this is such an important issue how can station data be left unreported?
Because Professor Phil Jones found that 100 well-placed stations around the world would be sufficient to determine hemispheric average trends. :)
Erudite_Eric wrote:
As for greenland, its the anomoly.
Right, so it's an anomaly. Now: Which station, period, degrees F or C, hotter than what?
If people made the effort to read something three times before commenting, blogs would be much more useful places. - Anon.
-
Erudite_Eric wrote:
Yeah, sure.
Not sure what you mean. What I proposed fits within the scientific methodology.
Erudite_Eric wrote:
No, what I did was look at how individual stations are sited and decide that adjusting for that using satellite images is insuficient. In stead urbanised stations should be rejected, and only rural ones kept in the data set.
That is a statement of an unproven theory. It seems provable but also must account for showing that once the data is massaged that way that it still represents a reasonable representation of actual weather. But until proven it does not refute the original link.
jschell wrote:
That is a statement of an unproven theory.
Theory? Go look at the pictures of badly sited weather stations in small towns in the middle of nowehere. Or dont, if you dont want to face reality.
============================== Nothing to say.
-
Erudite_Eric wrote:
Who was it who finally squeezed Hansens data and methods out of him a few years back
Strangely, the data processing method has always been available. However, the NCDC data were formatted for processing, i.e. binary data. The resultant GISS data were also held as binary files. I do remember reading the squeals when the binary files and FORTRAN programs were made available to the 'skeptic' community. (Huh! Amateurs! :-D ) Funny how the Open Source guys rewriting the FORTRAN code in Python and creating text data files had no problems communicating with GISS. Perhaps it was because their attitude was reasonable ( ;P ) rather than confrontational.
Erudite_Eric wrote:
a massive error in his work that actually ended up with the 30's being hotter than the 90's
No error in his work, rather a massive error in the nature of the US station data being passed to GISS by NCDC between 2000 and 2007. This affected US temperature anomalies from 2000. (Nothing to do with the 30's and 90's.) When the data were corrected in 2007, the error was shown to have had no effect on the Global Land-Ocean anomalies.
Erudite_Eric wrote:
As for reporting stations, I am not sure I believe you.
E.G., Canada. GHCN station count: 700-800 in the 1960's, 500-600 in the 1980's, -400 for 1990, <100 for 1991, ~50 thereafter. Although there are 1480 stations in Canada, only 49 are reporting to GHCN via the automated CLIMAT system. So the drop-out appears to be due to automated data collection.
Erudite_Eric wrote:
If this is such an important issue how can station data be left unreported?
Because Professor Phil Jones found that 100 well-placed stations around the world would be sufficient to determine hemispheric average trends. :)
Erudite_Eric wrote:
As for greenland, its the anomoly.
Right, so it's an anomaly. Now: Which station, period, degrees F or C, hotter than what?
If people made the effort to read something three times before commenting, blogs would be much more useful places. - Anon.
ict558 wrote:
is affected US temperature anomalies from 2000
It also reduced 1998 from hottest year to not so hot as the 30's. CLIMAT is not necessarially an automated system though, it is just a .txt format and data can be emailed, so why are less and less stations sending in data? And why does the reduction in the number of reporting stations coincide with an increase in temperature?
============================== Nothing to say.
-
jschell wrote:
That is a statement of an unproven theory.
Theory? Go look at the pictures of badly sited weather stations in small towns in the middle of nowehere. Or dont, if you dont want to face reality.
============================== Nothing to say.
Erudite_Eric wrote:
Theory? Go look at the pictures of badly sited weather stations in small towns in the middle of nowehere.
Or dont, if you dont want to face reality.The reality is that you haven't done the analysis that demonstrates that your theory significantly impacts the picture presented by the overall data. And of course if stations didn't exist in the "middle of nowhere" then that would be a problem because one needs a wide area of collection. So no idea what you think that using that denigrating statement is supposed to mean.
-
ict558 wrote:
is affected US temperature anomalies from 2000
It also reduced 1998 from hottest year to not so hot as the 30's. CLIMAT is not necessarially an automated system though, it is just a .txt format and data can be emailed, so why are less and less stations sending in data? And why does the reduction in the number of reporting stations coincide with an increase in temperature?
============================== Nothing to say.
Erudite_Eric wrote:
It also reduced 1998 from hottest year to not so hot as the 30's.
But not in any statistically significant way.
Erudite_Eric wrote:
CLIMAT is not necessarially an automated system though, it is just a .txt format and data can be emailed, so why are less and less stations sending in data?
NCDC do receive emailed corrections and station data. But the bulk of station data are transmitted as CLIMAT messages. I would imagine, in this era of austerity, that the reduction of personnel favours either automated or accessible stations. But that is totally without any research.
Erudite_Eric wrote:
And why does the reduction in the number of reporting stations coincide with an increase in temperature?
And improved GCE 'A' Level results? This has been addressed. A GISS analysis for 1880 - 2009 using only those stations reporting in 2009 is surprisingly close to one using all stations.
If people made the effort to read something three times before commenting, blogs would be much more useful places. - Anon.
-
Erudite_Eric wrote:
Theory? Go look at the pictures of badly sited weather stations in small towns in the middle of nowehere.
Or dont, if you dont want to face reality.The reality is that you haven't done the analysis that demonstrates that your theory significantly impacts the picture presented by the overall data. And of course if stations didn't exist in the "middle of nowhere" then that would be a problem because one needs a wide area of collection. So no idea what you think that using that denigrating statement is supposed to mean.
jschell wrote:
significantly impacts
We are talking about very small increases, increases inside instrumentation error bounds in some cases. What I meant with 'in the middle of nowhere' are small towns well away from large areas of night time lighting that have a temperature sensor sited near a heat source such as an air con heat exchanger, a tarmac parking lot, in the shade of a building, near a concreted area and so on. All these factors will affect the very cmall changes, fractions of a degree, which together from all station make up the overall picture. Adjusting for urbanisaiton using night time satelite images is just not specific and accurate enough. If you look at surfacestations.org you will get a clear idea of how badly some stations are sited. Therefore the proof is clear to the naked eye. There have been a number of studies of the efects of UHI and comparative studies of rural vs urban stations, and studies of temperature change vs population growth. They make interesting reading and show a lot more detail beyond the almost consistent 'global rise in temperature' shown in the major data sets. For exampler large parts of the far northern hemisphere but including the US were noticably warmer in the 1930s than of late. The vast majority of antarctica has cooled slightly and large areas of the tropics have not warmed significantly. (All this over the instrumental period of course). I am not denying that the planet has warmed since the little ice age, but that the warming is not so consistent and global as the rise in CO2 which is distributed fairly evenly around the world, so clearly there are some powerful regional factors at play.
============================== Nothing to say.
-
Erudite_Eric wrote:
It also reduced 1998 from hottest year to not so hot as the 30's.
But not in any statistically significant way.
Erudite_Eric wrote:
CLIMAT is not necessarially an automated system though, it is just a .txt format and data can be emailed, so why are less and less stations sending in data?
NCDC do receive emailed corrections and station data. But the bulk of station data are transmitted as CLIMAT messages. I would imagine, in this era of austerity, that the reduction of personnel favours either automated or accessible stations. But that is totally without any research.
Erudite_Eric wrote:
And why does the reduction in the number of reporting stations coincide with an increase in temperature?
And improved GCE 'A' Level results? This has been addressed. A GISS analysis for 1880 - 2009 using only those stations reporting in 2009 is surprisingly close to one using all stations.
If people made the effort to read something three times before commenting, blogs would be much more useful places. - Anon.
ict558 wrote:
But not in any statistically significant way.
Of course part of the problem is that the increases being discussed are so small that they lie within instrumentation errors, and out side stastical significance, regardless of which is the warmer or not. CLIMAT data can be sent by emil though. The automation you speak of is merely the format of the data. As for austerity, the station drop out started in the 1980s, no austerity then eh? :)
ict558 wrote:
A GISS analysis for 1880 - 2009
Well it would, asking the student to mark their own papers and all that. Of course, if the stations they have kept are the ones showing the greatest warming (why are their more airports as a percentage of sations today for example) then the rise will be seen. If on the other hand only rural stations are considered a different picture emerges (in the US, which has a large and long running data series. It is harder to do these kinds of comparisons globally due to the lack of coverage and questionable quality of data). Urbanised stations show an increase not seen in rural ones.
============================== Nothing to say.
-
jschell wrote:
significantly impacts
We are talking about very small increases, increases inside instrumentation error bounds in some cases. What I meant with 'in the middle of nowhere' are small towns well away from large areas of night time lighting that have a temperature sensor sited near a heat source such as an air con heat exchanger, a tarmac parking lot, in the shade of a building, near a concreted area and so on. All these factors will affect the very cmall changes, fractions of a degree, which together from all station make up the overall picture. Adjusting for urbanisaiton using night time satelite images is just not specific and accurate enough. If you look at surfacestations.org you will get a clear idea of how badly some stations are sited. Therefore the proof is clear to the naked eye. There have been a number of studies of the efects of UHI and comparative studies of rural vs urban stations, and studies of temperature change vs population growth. They make interesting reading and show a lot more detail beyond the almost consistent 'global rise in temperature' shown in the major data sets. For exampler large parts of the far northern hemisphere but including the US were noticably warmer in the 1930s than of late. The vast majority of antarctica has cooled slightly and large areas of the tropics have not warmed significantly. (All this over the instrumental period of course). I am not denying that the planet has warmed since the little ice age, but that the warming is not so consistent and global as the rise in CO2 which is distributed fairly evenly around the world, so clearly there are some powerful regional factors at play.
============================== Nothing to say.
Erudite_Eric wrote:
We are talking about very small increases, increases inside instrumentation error bounds in some cases...
Same problem with all of that as with the original link and posting. Your conclusions might be true. I never said otherwise. But you are not demonstrating that it isn't true. All your are doing is claiming it isn't and then presenting theories why it might not be true. And the link is worse because it does nothing but claim it is false. It doesn't even attempt to rationalize the stated conclusion (and as I previously suggested it leads one to suppose that the author doesn't understand data analysis.) There is a world of difference between claiming that the data is bad and showing that is bad.
-
ict558 wrote:
But not in any statistically significant way.
Of course part of the problem is that the increases being discussed are so small that they lie within instrumentation errors, and out side stastical significance, regardless of which is the warmer or not. CLIMAT data can be sent by emil though. The automation you speak of is merely the format of the data. As for austerity, the station drop out started in the 1980s, no austerity then eh? :)
ict558 wrote:
A GISS analysis for 1880 - 2009
Well it would, asking the student to mark their own papers and all that. Of course, if the stations they have kept are the ones showing the greatest warming (why are their more airports as a percentage of sations today for example) then the rise will be seen. If on the other hand only rural stations are considered a different picture emerges (in the US, which has a large and long running data series. It is harder to do these kinds of comparisons globally due to the lack of coverage and questionable quality of data). Urbanised stations show an increase not seen in rural ones.
============================== Nothing to say.
Erudite_Eric wrote:
The automation you speak of is merely the format of the data.
True, which (unless truly automated) would require some form of data entry. Hence, susceptible to reductions in personnel, as are inaccessible (non-automated) stations.
Erudite_Eric wrote:
As for austerity, the station drop out started in the 1980s
Hmmm.
Erudite_Eric wrote:
Over the last 20 years less and less stations are used to compose the data set, favouring lower altitude and lower lattitude stations and stations in urbanised environments, such as airports.
2011 - 20 = 1991, which would make you right (for once ;P ). From 1980 to 1991: Gentle decline. From 1992 to 1992: Precipitous drop.
Erudite_Eric wrote:
Well it would, asking the student to mark their own papers and all that.
An inept analogy. The GISS data and processes can be run by those other than GISS, no point in cheating. Similar exercises have been undertaken on independently modelled versions of the GISS methodology. By splitting the GISS data into two sets - those stations continuing to report since 1992, and those that have not reported since 1992 - each produced their own GISSTemp plot. The open source Clear Climate Code[^] lot, modelled in Python. Zeke Hausfather[^], modelled in STATA (2nd Chart). Neither shows any warming effect arising from the dropout. Anyhow, Watts and D'Aleo changed their minds about altitudes and latitudes: SURFACE TEMPERATURE RECORDS: POLICY DRIVEN DECEPTION? SUMMARY FOR POLICYMAKERS January, 2010: 5. There has been a severe bias towards removing higher-altitude, higher-latitude, and rural stations, leading to a further serious overstatement of warming. June 2010: 5. There has been a significant increase in the number of missing months with 40% of the GHCN stations reporting at least one missing month. This